Interventions early in the course of the illness could have an enormous impact on the public health costs and functional disability associated with schizophrenia. Early intervention requires a reliable method to identify young people who are at risk for developing psychosis. Youths with sub-psychotic or prodromal symptoms severe enough to prompt a clinical referral have a relatively high rate of conversion to psychosis. They often also exhibit subtle neurobiological and behavioral abnormalities, so-called endophenotypes, that are similar to those found in schizophrenia. These are thought to reflect the abnormal brain development that predisposes towards the illness. A judicious assessment of endophenotypes could, potentially, enhance the identification of individuals who are truly at risk for developing psychosis. To the extent that these developmental markers are evident before the impairment associated with sub-psychotic symptoms becomes obvious, they may also facilitate identification of at-risk youths at an earlier stage in the course of the prodrome. Research focusing on unaffected first-degree relatives of schizophrenia patients has identified many robust endophenotypes that are heritable, trait-like, and independent of overt illness. But these are also extremely poor predictors of future illness. This project will examine a set of neurobiological and behavioral markers that are likely to have more specific utility in predicting schizophrenia risk status. The rationale for this is based on two principles: 1)useful predictive measures should distinguish individuals with sub-psychotic clinical symptoms from unaffected first degree relatives without prodromal symptoms, who are unlikely to develop the illness~ 2) measures that can be mechanistically linked to a dysregulation of embryonic development during the critical late first - early second trimester risk period of fetal development should be more informative than relatively nonspecific indicators of compromised brain structure or function. It is proposed that measures of the neuronal integrity of the olfactory system represent just such specific risk markers. The olfactory system develops during this critical fetal risk perid in close coordination with development of the medial forebrain. It is intimately associated with the limbic brain regions commonly implicated in schizophrenia and is, itself, markedly impaired in schizophrenia. It can therefore illuminate early developmental contributions to heightened disease vulnerability. 35 patients with recent onset schizophrenia, 35 youths with a schizophrenia family member (genetic risk), 35 youths with prodromal symptoms (clinical risk), and 35 healthy comparison youths (low risk) will be studied. Behavioral tests will assess the integrity of olfactory capacity. 3T MR imaging and acoustic rhinometry will assess the structural integrity of the primary olfactory system. 7T 1HMR spectroscopy will assess the neuronal integrity of the primary olfactory system on a molecular level, through measurement of NAA and glutamate. The principal hypothesis is that schizophrenia patients and clinical-risk youths will both exhibit compromised olfactory neuronal integrity that distinguishes them from youths with isolated genetic risk or no risk for the illness.